Taylor's (Anti-) Epistemology

2

Taylor’s (Anti-) Epistemology HUBERT L. DREYFUS

INTRODUCTION Epistemology, as Charles Taylor understands it, is a discipline that arises along with the subject/object ontology introduced by Descartes. This ontology understands the subject as a self-sufficient mind related to the objects in the world by way of internal mental states that in some way represent those objects but in no essential way depend on them. The radical gap between what is inside the mind and what is outside in the world must be mediated in order for a subject to have knowledge of the world, and epistemology is the study of this mediation. In opposition to this Cartesian picture, Taylor describes the positive role our bodily skills and taken-for-granted background practices play in making sense of the world and in putting us in direct touch with everyday reality. But, at the same time, he stresses the negative role our modern taken-for-granted background framework plays in blinding philosophers to these phenomena. This blindness is characteristic not only of earlier versions of epistemology such as sense data theory, Kant’s scheme-content analytic, and Husserl’s phenomenological account of the mediational role of intentional content; it also casts doubt, Taylor seeks to show, on the claims of contemporary thinkers such as Donald Davidson and Richard Rorty, to have overcome epistemology. According to Taylor, these philosophers are still thinking within the inner/outer picture of our epistemic situation. In their version of it, we have access to the meaningful world and to the physical universe only insofar as it causally impinges on our sense organs and, thereby, produces our beliefs. In discussing Taylor’s anti-epistemology, I lay out, defend, and show the current relevance of his persuasive account of our direct encounter with the things in the world, and how this encounter grounds our knowledge. I then consider two challenges to this view: (1) The possibility that we are brains in vats, fooled by an evil computer scientist into thinking we are in direct

52

Taylor’s (Anti-) Epistemology

53

contact with the real world, whereas, in fact, there is no such world and (2) Rorty’s implicit critique that Taylor’s account of our embedding in the everyday world introduces a new version of the inner/outer distinction that undermines Taylor’s realist claim that natural science, at least in principle, is able to get outside our everyday experience and describe the physical universe as it is in itself. In response to these challenges, I argue that Taylor could accept the brain in the vat as a possibility, and, nonetheless, defend his basic antiepistemological argument. Second, I suggest that Taylor’s realism regarding the everyday world, far from standing in the way of scientific realism, enables him to counter Rorty’s antirealism with a robust realism concerning the entities described by physical theory.

OVERCOMING THE MEDIATIONAL PICTURE Taylor argues that no mental representations, be they sense data, visual experiences, or intentional content, and so forth, mediate our relation to everyday reality. The most recent and general version of the view Taylor opposes is found in Husserl’s Cartesian Meditations.1 As Husserl saw, and John Searle states clearly, all propositional intentional content is inner in the sense that the conditions of satisfaction formulated in the content depend on the mind and are independent of whether anything in the world satisfies them. In response, Taylor boldly states his thesis There is a big mistake operating in our culture, a (mis)understanding of what it is to know, which has had dire effects on both theory and practice in a host of domains. To sum it up in a pithy formula, we might say that we (mis)understand knowledge as “mediational”. In its original form, this emerged in the idea that we grasp external reality through internal representations. Descartes in one of his letters, declared himself “assur´e que je ne puis avoir aucune connaissance de ce qui est hors de moi, que par l’entremise des id´ees que j’ai eu en moi”.2 When states of mind correctly and reliably represent what is out there, there is knowledge.3

Taylor’s goal is to reveal the inner/outer structure of all epistemologies, even recent would-be anti-epistemologies, and to present and defend an opposed view, a view that denies that the inner/outer dichotomy in any form correctly describes our basic relation to reality. In response to the reigning mediational view, he draws on and elaborates Heidegger’s

54

Hubert L. Dreyfus

phenomenology of being-in-the-world, Merleau-Ponty’s account of our bodily being-devoted-to-the-world (ˆetre au monde), and Samuel Todes’s detailed description of how our body’s structure and its capacity for selfmovement structures the everyday world.4 These thinkers argue that nothing – not even propositional content – mediates our relation to everyday reality; that, at a level of involvement more basic than belief, we are directly at grips with the things and people that make up our world. Taylor elaborates an account of our direct interaction with the world in contrast to the mediational view: He notes that my ability to get around in this city or this house comes out only in getting around in this city or house. This important observation holds for the most global skills as well as the most local ones. Globally, it could be said, and, indeed, was said by Wittgenstein and Heidegger, concerning my ability to find my way about in the world.5 Taylor illustrates the antimediational point with a football: We can draw a neat line between my picture of an object and that object, but not between my dealing with the object and that object. It may make sense to ask us to focus on what we believe about something, say a football, even in the absence of that thing; but when it comes to playing football, the corresponding suggestion would be absurd. The actions involved in the game can’t be done without the object; they include the object.6

More locally, one might add, I can’t go through the motions of tying my shoelaces without holding on to the item in question, and I can’t tell which finger I use to type the letter “e” except by typing it. In general, unlike mental content, which can exist independently of its referent, my coping abilities cannot be actualized or, often, even entertained in the absence of what I am coping with. This is not to say that we can’t be mistaken. It’s hard to see how I could succeed in getting around in a city or in tying a shoe without the existence of the city or the lace, but I could be mistaken for a while, and, in the light of my failure to cope successfully, I may have to retroactively cross off what I seemingly encountered and replace it with a new understanding that amounts to directly encountering something else. Taylor’s originality and importance consist not only in his combining and drawing out the consequences of the best phenomenological accounts of involved coping; they also consist in his controversial claim that the mediational view he is attacking is still taken for granted by those who, in opposition to functionalism, defend qualia, and even those, like Donald Davidson and Richard Rorty, who claim to oppose all forms of subject/object, mind/world dichotomies. Taylor claims that these thinkers

Taylor’s (Anti-) Epistemology

55

are still imprisoned unawares in Descartes’ inner/outer picture. He points out that [Although] Quine denies Cartesian dualism by denying one of its terms – there is no “mental substance”, everything is matter, and thinking itself arises out of matter – [he] recreates a similar structure in the new metaphysical context. Our knowledge comes to us through “surface irritations”, the points in our receptors where the various stimuli from the environment impinge. Alternatively, he sometimes takes the immediate description of what is impinging, observation sentences, as basic, and he sees the edifice of science as built under the requirement that shows how (most of) these hold. In either variant, there is a mediational, or “only through” structure here.

Taylor goes on to show that those who at first sight seem to be opposed to Quine still hold that our knowledge of reality is necessarily mediated by propositional representations such as beliefs. He notes that Davidson quotes approvingly Rorty’s claim that nothing counts as justification unless by reference to what we already accept, and there is no way to get outside our beliefs and language so as to find some test other than coherence.7 And Davidson adds, “what distinguishes a coherence theory is simply the claim that nothing can count as a reason for holding a belief except another belief.”8 Indeed, Davidson explicitly “rejects as unintelligible the request for a ground or source of justification of another ilk [than belief].”9 An assurance that one’s view is self-evident, Taylor notes, is characteristic of what Wittgenstein calls a picture – a largely unreflected background understanding, that is, a way of seeing things, that seems so obvious, so commonsensical, as to be unchallengeable. That Davidson and Rorty are stuck in the Cartesian picture, Taylor claims, is evident in the way they assume, without argument, that the only way to ground a belief is to justify it rationally, on the basis of another belief. To help us appreciate what Davidson and Rorty miss, Taylor elaborates an account of our basic, preconceptual way of being in the world. [T]hings figure for us in their meaning or relevance for our purposes, desires, activities. As I navigate my way along the path up the hill, my mind totally absorbed anticipating the difficult conversation I’m going to have at my destination, I treat the different features of the terrain as obstacles, supports, openings, invitations to tread more warily, or run freely, etc. Even when I’m not thinking of them these things have those relevances for me; I know my way about among them. This is non-conceptual; or put another way, language isn’t playing any direct role.10

56

Hubert L. Dreyfus

Taylor explains in a personal communication that These relevances, which J. J. Gibson calls affordances – that the ground affords walking, water affords drinking, holes afford hiding, and so forth – are clearly meaningful relative to one’s interests and the structure of one’s body, yet they do not have to be experienced conceptually, i.e. our response to them need not be based on beliefs. We can, on reflection, note that boulders are obstacles, but we can just as well respond to their current relevance like non-linguistic animals.

On the basis of his description of our nonconceptual coping, Taylor objects to the Davidson/Rorty claim that knowledge consists solely of beliefs that are justified by other beliefs. He responds by contrasting this conceptual, and thus mediational, picture with a convincing description of how we acquire perceptual knowledge that is, how our perception-based beliefs are formed and come to be relied on. To show how the coherentist claim that reasoning from other beliefs is the only way particular beliefs can be grounded is so far from obvious as to be plain false, we need to step outside the mediational picture, and think in terms of the kind of embedded knowing which Heidegger, Merleau-Ponty, and Todes have thematized. Of course, we check our claims against reality. “Johnny go into the room and tell me whether the picture is crooked”. Johnny emerges from the room with a view of the matter, but checking isn’t comparing the problematized belief with his belief about the matter; checking is forming a belief about the matter, in this case by going and looking. What is assumed when we give the order is that Johnny knows, as most of us do, how to form a reliable view of this kind of matter. He knows how to go and stand at the appropriate distance and in the right orientation, to get what Merleau-Ponty calls a maximal grip on the object. What justifies11 Johnny’s belief is his being able to deal with objects in this way, which is, of course, inseparable from the other ways he is able to use them, manipulate, get around among them, etc. When he goes and checks he uses this multiple ability to cope, and his sense of his ability to cope gives him confidence in his judgment as he reports it to us.

Here Taylor is at his best. He shows that a description of our direct involvement with things is a convincing phenomenological answer to the dogmatic claim that the mind’s relation to the world must be mediated by beliefs caused by the things in the world. Perception provides reliable prepropositional bases for action and for accepting beliefs.

Taylor’s (Anti-) Epistemology

57

HOW TAYLOR’S PHENOMENOLOGY SUPPORTS AND EXTENDS THE WORK OF JOHN MCDOWELL Taylor notes the relevance of his phenomenological descriptions of everyday coping to the powerful critique of dualist epistemology mounted by John McDowell.12 McDowell is struggling directly with the inner/outer split implicit in Davidson’s and Rorty’s account of knowledge. He argues that beliefs alone cannot have content and connect with the world, and that only perception supplies the content of beliefs about reality. He therefore attacks the sharp demarcation between the space of reasons and the space of causes.

Taylor’s anti-epistemology has no place for this boundary either. His account of the formation of beliefs is meant to explain, as is McDowell’s, “how it can be that the places at which our view is shaped by the world, in perception, are not just causal impingings, but are sites of the persuasive acquisition of belief.”13 To begin with Taylor notes that: McDowell acknowledges that our perceptually formed beliefs are not just there as brute givens. Perception is precisely the activity whereby we have and can acquire more insight into why we have the beliefs we do. As McDowell says, the inclination “to apply some concept in judgment . . . does not just inexplicably set in. If one does make a judgment, it is wrung from one by the experience, which serves as one’s reason for the judgment. In a picture in which all there is behind the judgment is a disposition to make it, the experience itself goes missing.”14

Taylor endorses McDowell’s view: “Here is a phenomenological truth; and it points up something essential in the logic of the justification of our empirical beliefs; they do not start from pure givens that we cannot get behind.” This was the message in Taylor’s argument centering around Johnny checking the picture. Thus, in his description of the perceptual grounding, Taylor is entirely in agreement with McDowell. McDowell assumes, however, that to enter the space of reasons, perception must not only motivate beliefs but must justify beliefs by counting, as reasons for holding them, and so must be conceptual.15 But, Taylor asks, if, as McDowell seems to suppose, conceptually articulated perceptions just pop into our mind, the way beliefs do for Davidson and Rorty, why should we trust some of our perceptions more than others? For example, McDowell’s account does not help us understand

58

Hubert L. Dreyfus

why Johnny’s belief that the picture is askew, based as it is on his getting a maximal grip on the scene, is more reliable than Johnny’s belief that the moon, on which he can get no such grip, is bigger on the horizon than at the zenith. Taylor, therefore, seeks to show how an account of the basic levels of perception and the epistemic skills involved in forming a belief would enable McDowell to understand that there are degrees of perceptual support beneath rational justification – that our propositionally formed beliefs can only arise on the basis of a more basic skillful contact with the world that is prepropositional and in part even preconceptual. Taylor agrees with McDowell that reasoning is an exercise of a normguided capacity; it is thus an exercise of spontaneity in us, or otherwise put, of freedom. He notes that here McDowell is endorsing Kant: “When Kant describes the understanding as a faculty of spontaneity, that reflects his view of the relation between reason and freedom; rational necessitation is not just compatible with freedom but constitutive of it. In a slogan, the space of reasons is the realm of freedom.”16 Taylor continues to paraphrase and quote McDowell: Once we see the emptiness of the myth of the Given, our problem is somehow to bring this free spontaneity together with constraint. In order to stop the oscillation between the need for grounding which generates the myth of the Given, and the debunking of this myth, which leaves us with the need unfulfilled, “we need to recognize that experiences themselves are states or occurrences that inextricably combine receptivity and spontaneity”;17 we have to be able “to speak of experience as openness to the lay-out of reality. Experience enables the lay-out of reality itself to exert a rational influence on what a subject thinks.”18

Taylor agrees with McDowell that if we want to see how constraint and spontaneity come together, we have to find this in perception. But Taylor insists that to do so, we have to bring out how our ability to form beliefs like “the picture is crooked” draw on preconceptual epistemic skills. In Taylor’s estimation, the difference between him and McDowell seems to come down to this. We both give a crucial place to spontaneity in our most basic contact with the world; but McDowell doesn’t envisage any spontaneity which is not the exercise of concepts; whereas, following Merleau-Ponty, I have been describing precisely such a sub-conceptual exercise of spontaneity in our original grasp of our world.

Taylor’s (Anti-) Epistemology

59

To make his point, Taylor transposes the basic holistic arguments drawn from Kant from their original register into the preconceptual: Kant developed the original holistic argument on which all previous deconstructions of mediationalism have drawn. This is the argument against the atomism of the input, which consists in showing that any particulate percept has to be related to the world in which it figures, that we have necessarily to relate bits of knowledge [Erkenntnisse] to their object [Gegenstand ]. McDowell makes this same point: “The object of experience is understood as integrated into a wider reality, a reality that is all embraceable in thought but not all available in this experience”.19

But McDowell understands this holism as based on the way our “conceptual capacities” operate, whereas for Taylor, this kind of holism already functions on the level of preconceptual experience. As evidence Taylor cites MerleauPonty’s description of the skilled football player: For the player in action the football field is . . . pervaded with lines of force (the “yard lines”; those which demarcate the “penalty area”) and articulated in sectors (for example, the “openings” between the adversaries) which call for a certain mode of action and which initiate and guide the action as if the player were unaware of it. The field itself is not given to him, but present as the immanent term of his practical intentions; the player becomes one with it and feels the direction of the “goal,” for example, just as immediately as the vertical and the horizontal planes of his own body.20

For Taylor this description provides a good example of preconceptual spontaneity. Kant and McDowell speak of “spontaneity”, because they see the knowing agent not just passively receiving impressions from the outside world, but actively construing her surroundings, making sense of them. This we certainly do by applying concepts much of the time. But clearly not all the time. The football player is actively “making sense” of the field before him, articulating it into sectors, impregnable zones, possible “openings” between adversaries, vectors of vulnerability where the other team can break through; all without benefit of concepts – the terms we’ve applied here are ours, not drawn from his vocabulary.

Taylor concludes: Spontaneity at all levels is guided by the goal of getting it right; being clearly “forced” to come to some conclusion is not its negation, but its highest fulfillment. The same intrinsic relation between spontaneity and

60

Hubert L. Dreyfus

necessity that we see in the Kantian moral sage, and the Polanyian scientist, is visible in the lowly football player. He too is straining every faculty to get an accurate take on the ever-changing lines of force in the field. But the medium here is not moral reflection or theoretical representation, but the behavioral affordances of attack and defense.

Thus Davidson and Rorty may well be right that it is obvious that only beliefs can be reasons for accepting other beliefs, and maybe they are right that only beliefs can justify other beliefs, but McDowell and Taylor are surely right that it is a mistake to think that this shows that our only direct relation to the world is a causal one. Insofar as Davidson and Rorty ignore the actual evidence given by perceiving and take for granted that knowledge consists only in beliefs being justified by other beliefs, they seem to be captured by a form of the inner/outer mediational picture in which our beliefs are cut off from the external world. As McDowell complains, for Davidson, our beliefs are left frictionlessly “spinning in a void.”21 In sum, McDowell sees that beliefs alone cannot have content and connect with the world; that perception is necessary to connect beliefs to reality. In assuming, however, that perception must have conceptual content in order also to connect to the spontaneity of the space of reasons, he passes over the basic epistemological skills that make the conceptual content of perception possible in the first place. If McDowell wants a full account of the relation of mind and world, including why some perceptions are more reliable than others, he needs to take on board phenomenologists such as Merleau-Ponty, Todes, and Taylor. Their account of the basic levels of perception would enable him to make the last step to an account of the reliability of our direct perceptual contact with everyday reality.

DOES THE BRAIN-IN-A-VAT FANTASY POSE A PROBLEM FOR THE CONTACT REALIST? So far Taylor’s argument against the mind/world, inner/outer, dichotomy is on solid phenomenological ground, but he would like to go further. He would like to use the phenomenon of being-in-the-world he so convincingly describes, to cast doubt on the Descartes-inspired argument that we might, nonetheless, be brains in vats. It looks as though neuroscience calls into question any phenomenological account of our unmediated contact with an independent reality. It seems plausible to suppose that, as long as the

Taylor’s (Anti-) Epistemology

61

impulses to and from the nervous system reproduce the complex feedback loop between the brain’s outgoing behavior-producing impulses and the incoming perceptual ones, the person whose brain was being so stimulated would have the false belief that he was directly coming to grips with the things in the world. To defend his view that, in coping, the agent always directly encounters the real world, Taylor thinks he has to keep open the possibility that the causal basis of being-in-the-world is not in the head, or in the whole organism, or even the organism plus a whole virtual world, but that it might turn out that the basis of our experience of coping must be in the nervous system plus the whole actual world. To meet this challenge, Taylor begins by repeating his well taken phenomenological critique of behaviorism and cognitivism.22 “The idea is deeply wrong that you can give a state description of the agent without any reference to his/her world (or a description of the world qua world without saying a lot about the agent).” But he then moves from the agent to the causal level: That’s why I find the brain-in-a-vat supposition so unconvincing. It seems to me to rely on the old Cartesian separation mind/world, and just to transfer the first term into a material register. But once you take account of the embedding of practice in body-world, the whole idea gets less convincing.

Taylor therefore questions the generally accepted belief that the experience of being-in-the-world, like any experience, must be, as Searle puts it, “caused by and realized in the nervous system.”23 He responds that “maybe the minimum system which can duplicate the experience of a human being in the world is a human being in a world.” But, granted that our skill for getting around in the world can only be experienced as functioning in direct contact with concrete situations, why should we need more than a virtual world to reproduce this experience? As Searle says, “[E]ach of us is precisely a brain in a vat; the vat is a skull and the ‘messages’ coming in are coming in by way of impacts on the nervous system.”24 Taylor’s response is that Experience supervenes on a whole human organism operating in its environment; that we know. Now maybe we can show that it supervenes on something less than this. But how can we know this? . . . Knowing that the bodily medium is crucial doesn’t tell us how it is, and therefore gives us no way of knowing how to identify bits which are crucial.

62

Hubert L. Dreyfus

Taylor is right that the Descartes/Searle picture is by no means as obvious as current philosophers seem to suppose. But as long as Taylor admits that determining what conscious experience supervenes on is an empirical question, we have to take seriously the (admittedly far-fetched) possibility that experience could supervene on a nervous system connected to a virtual world – that, therefore, we may be brains in vats whose experiences are produced and coordinated by an intelligent computer, as in the movie The Matrix.25 If this scenario is even a wildly remote empirical possibility, indeed, even if is understood only to be an intelligible possibility, it appears that the phenomenology of direct coping cannot be used to counter the epistemological claim that all of our experience of the world is indirect.26 This seeming conflict between neuroscience and phenomenology arises because, in Taylor’s description of perception as unmediated contact with reality, he claims too much. Remember that, according to him, we can’t get around in a house, a city, or the world without interacting with that house or that city. But in the brain-in-the-vat fantasy, there is no house and no city, indeed, no real world, to interact with. It seems that it is not strictly true that, as Taylor likes to quote Merleau-Ponty, “To ask if the world is real is to fail to understand what one is asking.”27 Once one has read Descartes or seen The Matrix, the question seems at least to make sense. But it seems to me Taylor should not worry. Whether an agent’s relation to the world is direct or mediated is a phenomenological question. It could not possibly be supported or refuted by an answer to the empirical question as to whether the processing that underlies the experience of being-in-theworld takes place in the brain, the nervous system, the organism, or the whole physical universe. All Taylor can claim, and all he needs to claim to defend his antimediationalist view, on the basis of phenomenology, is that even in the case of the brain in the vat, the people whose brains are getting virtual reality inputs correlated with their action outputs are directly coping with perceived reality. Even in The Matrix world, people play football with footballs, relate to chairs by sitting on them, and find their way around in their world without representing it in their mind. The important point is that, even in the world of a brain in a vat, coping is more direct than allowed in any of the mediational views that have been held from Descartes to Rorty. On these views, the content of our beliefs can be entertained without taking a stand on the existence of the objects that would make these beliefs true. But, even if the brain-in-the-vat fantasy makes sense, Taylor can still hold the radically antimediationalist view that, no matter whether the brain is in a cranial vat in an organism coping with the world, or a ceramic vat interacting with a computer, “[t]he very idea of

Taylor’s (Anti-) Epistemology

63

an inner zone with an external boundary can’t get started here, because our living things in a certain relevance can’t be situated ‘within’ the agent; it is in the interaction itself.”28 So he should still conclude that, even in The Matrix world, “the idea is deeply wrong that you can give a state description of the agent without any reference to his/her world (or a description of the world qua world without saying a lot about the agent).” Another way to make this point is that one should sharply distinguish the skeptical/epistemological problem as to whether our inner mental states accurately represent what is out there in the external world from a phenomenological description of an agent’s relationship to the world. The issue for Taylor should not be whether the world is as we believe it to be. That is Cartesian doubt. The phenomenological point is that our direct contact with the perceptual world is more basic than belief. So when MerleauPonty says that to ask if the world is real is to fail to understand what one is asking, he must mean that, even in The Matrix world, at the basic level of involved skillful coping, people would still be “empty heads turned towards one single self-evident world where everything takes place.”29 Thus, Taylor’s phenomenological account of being-in-the-world and his consequent critique of mediationalism is untouched by the conceivability of our being a brain in a vat.30

THE REAL PROBLEM: ANTI-EPISTEMOLOGY AS ANTIREALISM Paradoxically, however, once we recognize that in perception we directly encounter everyday perceptual objects on the background of our unmediated embedding in the everyday perceptual world, it looks as though we can no longer make sense of the idea that we are capable of knowing things as they are in themselves, that is, independently of the way they make sense to us in our embodied interaction with them. It seems that, at the primordial, preconceptual level of perception and action, we are confined to the cross-cultural clearing opened up by our ability to respond to affordances, and, at a higher level, we are imprisoned in the general style of our culture’s coping practices. Indeed, just insofar as our everyday coping practices give us direct access to our world, they seem to block access to, indeed, make unintelligible, the very idea of access to the universe as it is in itself. Rorty is happy to embrace this consequence. He holds that we are confined to what can be encountered on the basis of our coping practices.31 We therefore shouldn’t think of science as a way of discovering propositions that correspond to an independent reality, and, fortunately, we don’t

64

Hubert L. Dreyfus

need to. Embedded coping is the only realism we can make sense of, and all the realism we need to make sense of science. Taylor, however, although following Merleau-Ponty in thinking of our relation to the everyday world as a “co-production,”32 also wishes to defend a robust realist view of science as giving us access to things as they are in themselves, independent of their relation to our bodily and cultural coping skills. Rorty highlights his opposition to Taylor on this point: Realism becomes interesting only when we supplement plain speech and common sense with the “in itself” versus “to us” distinction. Taylor . . . thinks that this latter distinction cannot simply be walked away from but must be dealt with. I think neither he nor anyone else has explained why we cannot just walk away from it. Such an explanation would have to tell us more than we have ever before been told about what good the distinction is supposed to do us. I keep hoping that Taylor, as fervent an anti-Cartesian as I, will join me in abandoning it. Alas, he persists in agreeing with Bernard Williams, . . . and other admirers of Descartes that it is indispensable.33

Here we have the parting of the ways between two views. On the one hand we have Rorty’s, which I shall call deflationary realism, that claims that the objects of science are only intelligible on the background of our embedded coping, so that the idea of a view from nowhere is literally unintelligible. On the other hand there is Taylor’s view, which I shall call robust realism, which claims that, to understand the status of the structures studied by natural science, we have to make sense of an absolutely independent reality. From the perspective of the robust realist, deflationary realism is a kind of antirealism that succumbs to a new inner/outer distinction. Rorty holds that stressing being-in-the-world blocks direct access to the universe. How can Taylor defend the seemingly contradictory claims that the primordial and unavoidable significances of things are or are connected to our bodily existence in the world, and that, nonetheless, we can make sense of a science of the components of the universe as they are in themselves, utterly independent of any relation to our embodiment? Taylor responds that: Our humanity consists in our ability to decenter ourselves from our original mode of absorbed coping; to learn to see things in a disengaged fashion, in universal terms, or from an alien or “higher” point of view. The peculiar form that this takes in Western scientific culture is the attempt to achieve, at least notionally, a “view from nowhere”, or to describe things from the “absolute standpoint”. Only we have to see that this decentered mode,

Taylor’s (Anti-) Epistemology

65

whatever form it takes, is in an important sense derivative. The absorbed, engaged one is prior and pervasive.

But such a response raises more questions than it answers. If the engaged experience is primordial and the disengaged mode is derivative from the engaged one, what sort of view from nowhere can we hope to achieve or even approach? It seems to follow, rather, that whatever we can encounter is a function of the kinds of bodies and needs we ineluctably have. If, for example, our body structures our experience of spatiality and temporality as Merleau-Ponty contends, Todes works out, and Taylor accepts, how can we prescind from our bodily-relative sense of reality and still have a science of the motions of objects in the spacio-temporal universe? How could Taylor’s view be anything but a refined variant of deflationary realism?34 Indeed, Taylor seems to make exactly the deflationary move he wants to resist when he says that: The mediational view provides the context in which the whole complex of issues around “realism” and “anti-realism” make sense. They lose this sense if you escape from this construal, as Heidegger and Merleau-Ponty have done. Or perhaps better put, one awakes to an unproblematic realism, no longer a daring philosophical “thesis”.

This sounds exactly like Rorty’s reading of Heidegger and Merleau-Ponty. Indeed, it looks as though Taylor’s “unproblematic realism” concerning the things that we ordinarily cope with – things that would not be the sort of spacio-temporal things they are independent of our embodied coping – when made the basis of a “realism” concerning the universe, is precisely deflationary realism. As Rorty puts it “Taylor thinks that once one gets out from under epistemology one comes to an ‘uncompromising realism.’ I think one comes to a position in which the only version of ‘realism’ one has left is the trivial, uninteresting, and commonsensical one which says that all true beliefs are true because things are as they are.”35 So, according to Rorty, there is nothing more one can say about what makes the propositions of science true than we can say about what makes the claims we make about baseballs true. They both report how things are, and they both depend on our embedding. It is, thus, not only false but unintelligible to hold, as Bernard Williams, Thomas Nagel, and Taylor do, that truths about baseballs depend on our bodies and our cultural agreements, whereas the truths of science describe things as they are totally independently of us and our everyday way of making sense of things.

66

Hubert L. Dreyfus

If all we can say about our scientific truth claims is they are no more nor less relative to our sense-giving capacities than our everyday truth claims, it’s hard to understand how Taylor’s account of our direct embodied encounter with everyday reality is supposed to make the idea of a view from nowhere even notionally intelligible. The emphasis on a background of intelligibility correlative with our bodily structure would seem to argue for just the opposite conclusion – a view from somewhere, namely, from within our embodied embedding. In comparison with robust realism’s thesis that science studies things as they are independently of us, unproblematic realism does not seem to be realism at all. Indeed, unproblematic realism with respect to the everyday world makes realism concerning the universe highly problematic. This is where Taylor makes an original move that diverges not only from Rorty but also, in different ways, from the existential phenomenologists he generally agrees with, and even from some of his own claims. He tells us that: If we see that our grasp of things is primordially one of bodily engagement with them, then we can see that we are in contact with the reality which surrounds us at a deeper level than any description or significance-attribution we might make of this reality, and that this dissolves the temptations to anti-realism.

That is, significance depends on our coping, but, in coping, we sense we are in touch with a reality more basic than significance. It remains to be seen, however, if this reality beneath significance is more than the brute causality independent of any description invoked by Rorty; whether it has the structure attributed to it by science. As we have seen, Taylor points out that even when I’m thinking about something else, my body takes account of obstacles such as boulders and the like without my needing to be aware of them at all. This gives us an ahistorical, cross-cultural commonality with all creatures that have bodies of roughly our size, shape, and power, no matter how, on reflection, they classify boulders – whether in their world boulders are sacred objects, lookout towers, or simply rocks that get in the way of climbing. But this still leaves unclear how a cross-cultural experience of obstruction, for example, could in any way justify the claim that, in responding to affordances, one is responding in a direct way to how the universe is in itself independent of all our significance attributions. No one could doubt that the significance of boulders as obstructions depends on our kinds of bodies and our kinds of desires.

Taylor’s (Anti-) Epistemology

67

So we still need to understand how, what is experienced in perception at a level deeper than significance, is the real as it is in itself. Taylor explains that “The most fundamental, rock-bottom feature of our general take on the world is that it surrounds us, gives us things, but also withholds, threatens to annihilate or hurt us sometimes, allows us to do some things, and resists others.” But again one wonders why “surrounding,” “giving,” “withholding,” “threatening,” and “resisting” are the way things are in themselves rather than our way of making sense of them. A better case for our getting in touch with the way things are in themselves is offered by Todes. He points out that our experience of having to balance in a gravitational field gives us the sense of a force independent of us that we have to conform to, a force which sustains our coping only if, by balancing, we relate to it on its terms.36 But again one can ask whether our relation to this force as something that pulls us down but also enables us to stand up, isn’t our way of making sense of nature in our world, rather than an experience of how the universe is in itself. In each of the above experiences we make sense of the experience as something that conditions us, sets boundary conditions on our ability to cope, and thereby reveals something outside our coping powers. But in so doing we have not yet arrived at a primordial bodily “contact with the reality which surrounds us at a deeper level than any description or significanceattribution we might make of this reality.” We have only arrived at the paradoxical significance attribution of something that somehow exceeds our significance attribution. This is certainly not enough to dissolve “the temptations to anti-realism.” It looks as though whatever is independent of us is so inextricably bound up with our coping capacities that it can only be understood as a boundary condition on our activity. But natural science doesn’t just run up against the boundaries of our coping, it claims to reveal the intelligible structure of the universe as it is in itself. If all that active perception gave us were a sense of something independent of us that set limits to our coping, we would not be able to describe the universe from nowhere, and we would be left with deflationary or internal realism – science at best describing what Merleau-Ponty calls “the in itself for us.”37 But Taylor’s novel approach only begins with the above account of our encounter with the sustaining and threatening phenomena that condition our coping. He turns from his description of the otherness revealed by all coping to an even-more-overlooked component of our primordial perceptual experience, namely, that, in skillful coping, as in balancing in the vertical field or climbing over boulders, our skill consists in getting in sync

68

Hubert L. Dreyfus

with the structure of the universe the better to cope with it. Consider what is involved in grasping and drinking from a glass that lies to hand. To begin with, I have to see the glass. This is no mean feat. To perceive at all, we have to align ourselves with the causal powers of the universe. As embodied beings, we have to face what we are looking at, move to an appropriate distance given the size of the object, and assure an unencumbered line of sight to it. In this way, our embodied skill spontaneously takes account of the fact that, as the causal theory of perception makes clear, in order to see an object, we have to be in a position for our eyes to be causally acted upon by light from it. Thus, the universe constrains us, and rewards us with sight only insofar as we conform to its causal structures. But we are so skilled at getting an optimal take on things that, unless there is some disturbance, we overlook the fact that we once had to learn to align ourselves with the constraints of nature in order to perceive. In general, the universe solicits us to get a better and better grip on its causal structure, and rewards us with more and more successful coping. Our coping skills thus put us in touch with the structure of the causal powers of nature, not just its brute impinging, thereby bridging the gap between the meaningless brute causal influence acknowledged by Rorty and our perception of a meaningful perceptual world accepted by McDowell. Taylor is thus able to specify the way in which perception gives us access to the causal structure of the universe, revealed in our way of making sense of things, but independent of our everyday significance attributions. One can only see this if one uses phenomenology to uncover the coping skills covered up by the conceptually permeated perceptual world where analysis normally starts. Only by accepting Rorty’s challenge that he tell more than we have been told before in the history of philosophy, is Taylor able to show us what philosophers close to common sense like Aristotle have suspected all along: that we experience ourselves as perceptually in touch with the cosmos. But we can only understand how we are normally in touch with things in themselves when we see that we are not disembodied, detached contemplators, but rather, embodied, involved coping agents.38

SUPERSESSION To understand Taylor’s robust realism vis-`a-vis science, we now need to extend his claim that in our everyday coping experience we sense that we

Taylor’s (Anti-) Epistemology

69

are in contact with a nature with a structure of its own that supports our coping, to the stronger claim that it makes sense to think we can correctly describe that structure and that, indeed, there is evidence that our current science may well be progressively getting it right about (at least some aspects of ) the universe. To do so, we need to introduce and defend Taylor’s account of supersession. Taylor begins by reminding us that when we confront anomalies in our perceptual experience, we know how to find out what’s wrong with our current understanding and improve it. If we enter a cafe that seems too large for the building it is in, we are confused until we resolve this anomaly by noting that the walls are covered by mirrors. Then things snap into place, and our confused, partial perceptual grasp becomes clear and secure. As we explore a city, we gain a more and more perspicuous understanding of it; we are no longer surprised and disoriented at each turn. In general, in our everyday perceptual encounters with the world, we are solicited to move toward an ever more clear and secure grasp of our surroundings. In understanding other cultures, this process of moving from confusion to clarity is much more difficult. Thanks to our shared embodiment, however, we can make some progress toward an understanding of what is going on in another culture by noticing that its members respond to many of the same affordances we do. But we may nonetheless find that their understanding of the sacred, to take Taylor’s favorite example, makes no sense to us. How could the Aztecs in a sacred service tear out and eat the heart of sacrificial victims, who, strangely, don’t even think of themselves as victims, but seem to be honored by thus being killed? Taylor points out that we cannot even be sure that “sacred,” “sacrifice,” and “honor” are the appropriate terms here. He tells us that in such cases: What is needed is not the Davidsonian “principle of charity,” which means: make the best sense of them in what we understand as sense; but rather: coming to understand that there is a very different way of understanding human life, the cosmos, the holy, etc. Somewhere along the line, you need some place in your ontology for something like “the Aztec way of seeing things”, in contrast to “our way of seeing things”; in short, something like the scheme/content distinction.

But this pluralism need not lead to antirealism. According to Taylor, we have a sense that we are open to something that is independent of any of our interpretations – something that sets limits on what takes on reality are livable. We can thus accept that there are radically different cultural

70

Hubert L. Dreyfus

understandings of being, and yet rank their relation to reality. Thus Taylor continues: We can see how the embedded view offers resources for recognizing differences of scheme, without generating arguments for non-realism. There may be (and obviously are) differences, alternative takes on and construals of reality, which may even be systematic and far-reaching. Some of these will be, all may be, wrong. But any such take or construal is within the context of a basic engagement with or understanding of the world. It is in virtue of this contact with a common world that we always have something to say to each other, something to point to in disputes about reality.

All the various cultural interpretations are for us, so what Taylor rightly considers realism with respect to cultural styles, looks like antirealism with respect to science. But it turns out that science has its own internally generated way of progressing. Rather than accepting with Heidegger and Thomas Kuhn that the worlds of Aristotle and Galileo can’t be compared because they were asking different questions, Taylor looks for the specific anomalies that Aristotelians ran up against that Galileo could account for – in this case that missiles from catapults and guns don’t head straight for their natural place at the center of the earth, but follow a parabolic path. Or, to take a simpler example, by assuming that the earth moves rather than the sun, Copernicus could give a more perspicuous account of the motion of the heavenly bodies than epicycle theorists could, and astronomers could then see that the sun could better be understood as a star than as a planet. Moreover, Taylor points out, even the cultural background understandings on which the methods of science are based can progress. He notes that Kepler’s success in getting a more complete and clear grasp of astronomical phenomena showed that the Aristotelian background understanding that one could not, and should not, try to account for all phenomena – both terrestrial and superlunary – in the same way, could be improved on. That undermined the Greek understanding of science as empiria, just as Galileo’s findings undermined the medieval idea of scientia, and eventually led to the modern understanding of science as research. As Heidegger points out, research differs from scientia and empiria in proposing a universal ground plan and then trying to fit all phenomena, even those that look like anomalies, into that plan, rather than just dismissing anomalies as unnatural events, monsters, or miracles.39 This new understanding of science as world-picturing, in turn, gave us a more coherent and powerful understanding of nature. More phenomena could be lawfully related according

Taylor’s (Anti-) Epistemology

71

to the Newtonian ground plan than by Aristotelian generalizations from everyday experience. In this way, according to Taylor, we can see that scientific revolutions are cases of supersession. If new overall conceptions are not immediately accepted it is because the older view is entrenched, not because the proposed changes are not rationally motivated. The proof that one view is superior to another, Taylor holds, is that once one understands the new overall way of looking at things in science, as in the case of the mirrors on the caf´e walls, there is no way of going back and accepting one’s former understanding. Thus, the direct coping that gets us in touch with a shared everyday world, and gives us a sense of an independent nature that sets limits to what we can do unless we get in sync with it, puts us on a path that leads to theories that correspond more and more adequately to the structure of the universe.

TAYLOR VERSUS RORTY ON TRUTH Rorty, however, thinks one can accept Taylor’s account of supersession in science, namely that science progresses by accounting for anomalies and thereby giving a more and more coherent account of the phenomena, without accepting Taylor’s claim that science thereby arrives at theories that correspond better to an independent reality. As Rorty puts it, “Believers in the correspondence theory have to claim that some vocabularies (e.g. Newton’s) do not just work better than others (e.g., Aristotle’s) but do so because they represent reality more adequately. Taylor thinks that good sense can be made of this claim, and I do not.”40 Taylor’s first response to Rorty on this point is simply to defend the correspondence theory of truth as our unproblematic commonsense understanding of the way we check our assertions against the facts. Lots of simple everyday sentences are meant to communicate the way things are; they give a “picture” of how things stand, and they are correct if the way things really stand corresponds to this picture. It is in this common and well-understood sense that many ordinary indicative sentences “represent” what they’re about: there are 15 chairs in this room. Well, are there really? – Count them.

This disarmingly simple example of counting the chairs in the room is full of problems, however, as Taylor, given his interest in Wittgenstein, surely recognizes. How many chairs there are in a given room depends on the interests of the one counting them. If I want to seat students right

72

Hubert L. Dreyfus

now, then I don’t want to count broken chairs and those in boxes waiting to be assembled. Or maybe I’ll count the broken chairs, too, if they are not so badly damaged that one can’t sit on them in a normal way. But, if I’m determining how many students the classroom will seat next semester, I’ll count the unassembled chairs, too (but still not the badly broken ones). Of course, I can always spell out the assertion whose truth I want to check so as to avoid this sort of contextual under-determination. “Johnny, how many currently usable chairs are there in the room?” But, as Wittgenstein points out, one can never build the whole background into the explicit proposition whose truth can be spelled out so that it is convincing. Truth amounts to correspondence between our description of the facts relative to our interests and the facts themselves. As a deflationary realist, Rorty is happy to take over the way Taylor’s realistic account of truth as simple correspondence seems to level the difference between truths about the things in the world like chairs that are relative to our background concerns, and truths about the stuff in the universe that allegedly are not. Rorty simply levels the difference in the opposite direction from Taylor and argues that, since the truth about chairs requires a shared commonsense background, so must the truth about neutrinos: Taylor seems to think that neither I nor any one else would feel any “serious temptation to deny that the no chairs claim [“There are no chairs in this room”] will be true or false in virtue of the way things are, or the nature of reality.” But I do, in fact, feel tempted to deny this. I do so because I see two ways of interpreting “in virtue of the way things are.” One is short for “in virtue of the way our current descriptions of things are used and the causal interactions we have with those things.” The other is short for “simply in virtue of the way things are, quite apart from how we describe them.” On the first interpretation, I think that true propositions about the presence of chairs, the existence of neutrinos, the desirability of respect for the dignity of our fellow human beings, and everything else are true “in virtue of the way things are.” On the second interpretation, I think that no proposition is true “in virtue of the way things are.”41

To resist Rorty’s leveling of the world and the universe, Taylor would have to highlight the difference between the way correspondence works in the everyday world and in science. He could point out that, unlike the number of chairs, how many neutrinos there are in a given space does not depend on how they fit with our everyday practices or what anyone, even the scientist who makes the claim, wants to do with them. The development of science shows that science gets better and better at what Heidegger calls

Taylor’s (Anti-) Epistemology

73

deworlding, and that, where science is concerned, supersession leads us from relatively contextually determined (worldly) truth claims to relatively context-free (deworlded) truth claims about features of the universe as it is in itself.42 (Not that we can ever be certain that the entities dealt with by our current science are fully deworlded.) It might have turned out that all we could know about the independent structure of the universe was relative to our background understanding of the boundary conditions it placed on our activity. But Taylor reminds us that Galileo and company discovered that we could bracket our direct, embodied experience of the everyday world, leaving out more and more of the significance of things that depends on our social practices; as well as prescinding (as Taylor puts it) from the properties of everyday things that depend on our senses and on the shape and capacities of our bodies. We are able thereby to discover and investigate a physical universe with no perceptible things with their colors, orientation, solidity, weight, and so forth, where there is no near and far, no up and down, and no earlier and later. Moreover, it happily turned out that this deworlding, as Heidegger calls it, was not merely a negative accomplishment, but that, when we left behind the world of everyday experience, we discovered universal causal laws, and natural kinds some of whose properties explained why the kinds in question were describable by just those causal laws. But to see that some theories don’t just “work better than others” in some general pragmatist way, but that we need to and can make sense of the claim that some theories fit the universe better than others, we again have to go back to our sense of embodied coping. Taylor notes that if we were to take our beliefs as given and warranted only by other beliefs and didn’t investigate how these beliefs were formed, Rorty would be right that a correspondence theory of truth adds nothing to what we already say when we affirm the truths of science. But this doesn’t mean that talk of correspondence doesn’t say anything important. If it adds nothing, it is because the understanding that our claims to truth are grounded in our epistemic skills for getting a grip on reality is an implicit part of the background understanding that underlies our pursuit of science.

Taylor points out that “the things that show up for us as obstacles, supports, facilitators, in short as affordances, have as it were an ontic solidity and depth.” They have what philosophy will later call their “nature,” which we have to respect and adjust ourselves to. Just insofar as coping requires this adaptation, and our actions confirm that we are getting a better and better

74

Hubert L. Dreyfus

grip on nature, correspondence makes sense to us as a way of describing our relation to reality. Only when our theories about nature have been arrived at step by superseding step, following the demand for clarity and control that was implicit in our original perceptual encounter with an independent reality, have we the right to claim correspondence for our theories. Approaching pure correspondence – the view from nowhere – means approaching the goal of getting a maximum grip on the universe, thus satisfying the call for a clairvoyant account of reality that, from the start, drew us toward an increasingly encompassing and refined grasp of the independent reality that supports our coping. To sum up, Taylor holds that our experience of our body’s engaged coping has four basic, interrelated, characteristics that ground intelligibility – both that of the everyday world and that of the universe as revealed by science. First, we experience ourselves as open to the real. As Taylor puts it: “The reality of contact with the real world is the inescapable fact of human (or animal) life, and can only be imagined away by erroneous philosophical argument.”43 Second, we experience whatever we encounter as having aspects that go beyond what we can grasp at present. This is not just a fact about the things we perceive but, as Merleau-Ponty points out, a fundamental structure of perception. Third, we are drawn to make sense of things, to move toward getting a more and more coherent, and encompassing grip on them – as Merleau-Ponty puts it, “we are condemned to meaning.”44 As we have seen, Taylor calls this progressive movement toward making more and more sense, not only of the everyday world but also of reality in general, supersession. Fourth, in the process of gaining a more and more clairvoyant grip on our everyday world, we have discovered that we can leave behind all body-relative properties such as accessibility, color, size, and even the sort of everyday spatiality and temporality that are intelligible only on the basis of our embodied coping. We thereby gain the possibility of a theoretical grasp of physical reality that is alien to our everyday embodied mode of perception of space, time, objects, and causality, and which, for that very reason, could claim to correspond to the universe as it is in itself. Note that Taylor is arguing against the deflationary realist and so only needs to claim that, given our coping experience, robust realism concerning science is intelligible – that it makes sense to think of science as attempting to describe the universe as it is in itself. He does not have to hold that our current science is getting it right about the universe or that any science will ever get it right; nor does he have to claim that we could ever know for sure whether the current science is in fact on the right track. These are

Taylor’s (Anti-) Epistemology

75

epistemological questions, whereas Taylor and Rorty are concerned with the ontological question as to the mode of being of the entities studied by science. Taylor go further, however. As we have just seen, he argues convincingly that, given our science’s supersession claims, it makes sense to hold that our science is in fact zeroing in on (one aspect of) the physical universe as it is in itself. Taylor goes beyond deflationary realism by showing how (1) our primordial embodied and embedded grasp of reality, far from standing in the way of robust realism, makes intelligible how our body, cultural history, and language give us access to a universe whose intelligibility in no way depends on any structures of our embodied way of being in the world; and (2) supersession supports the view that we may well be learning more and more about the causal structure of that universe as it is in itself.

TAYLOR’S PLURALIST ROBUST REALISM Still, deflationary realists such as Rorty can persist. It looks as though no description of our direct embodied encounter with everyday reality, even if that reality is experienced as independent and inexhaustible, could make the idea of a view from nowhere intelligible. Taylor’s emphasis on a background of intelligibility correlative with our bodily structure would seem to argue for just the opposite – a view from within our embodied embedding. Rorty contends that, granted we experience the causal character of the boundary conditions to which we must conform, how the structure of these boundary conditions is to be described must always be relative to our vocabulary, practices, and bodily coping capacities. Thus Rorty thinks he has the right to reject Taylor’s realism on the grounds that the idea of a correct description that corresponds to the structure of the universe as it is in itself makes no sense. His actual argument, however, consists in ridiculing the idea that the universe has a language of its own. And, of course, the universe doesn’t speak. But, all the same, the universe could have a structure whose essential properties were definable in a language it led scientists to adopt. But how could the universe possibly have its own proper description? And how could it lead us to better and better ones? Taylor answers that it could if there were natural kinds with essential properties, and if we could designate the kinds we encounter by means of a provisional mode of reference using everyday language that remained noncommittal as to which, if any, of the descriptions we use to refer to the kind in question were

76

Hubert L. Dreyfus

essential to it. Both Heidegger and Saul Kripke have defended versions of this idea.45 So, to take two of Kripke’s well-known examples, we could start by investigating some shiny gold-colored stuff and eventually find out that its essence is to have an atomic number of 79, regardless of whether or not it is gold-colored. Or we could provisionally identify lightning as a flash of light in the night sky and eventually find out that it is essentially an electrical discharge even when we can’t see it. Thus, a purported natural kind is first designated by a description that points out an instance of it – that yellow stuff. This pointing fixes the reference but does not commit the designator to the claim that the description used in pointing out the kind has grasped the kind’s essential property – the property that causally explains all the other physical properties. Thus, the initial description, although relative to our everyday language, interests, and capacities, leaves open the possibility that investigation may discover the kind’s essential properties. In this way, Kripke shows that the way demonstrative reference works makes intelligible the idea of an access to natural kinds (if there are any) whose essential properties we can describe in a language appropriate to them. Rorty would be quick to point out, however, that the question of the relativity of descriptions reappears in our understanding of what counts as an essential property. Must an essential property of a kind explain all and only the causal properties interesting to physics? If so, the essential property of gold may well, indeed, be having an atomic number of 79. But Taylor, with his openness to other cultures having an understanding of nature different from ours, would not want to accept this Kripkean view, since it follows from it that most of the beliefs about nature held by people in other cultures are false. Take gold, for example. For our science, its essential property is the one that explains how it falls under a large number of universal causal laws that explain its ductility, conductivity, malleability, solubility, capacity to form certain compounds and not others, and so forth. But for another culture, say the Ancient Egyptians, gold’s essential property might have been that it was sacred and so shone with divine radiance. How can Taylor claim that true scientific assertions pick out the essential properties of things as they are in themselves, without accepting the implication of Kripke’s scientific realism, that, insofar as our scientific understanding of nature is true, the understanding of nature of cultures that don’t share our understanding must be false, that, for example, the Ancient Egyptian understanding that the essential property of gold is the sacred powers that cause its radiance is simply mistaken?

Taylor’s (Anti-) Epistemology

77

To begin with, we have to remember that ours may well be the only culture that claims that, if true, our theories concerning the kinds of entities in the universe correspond to those kinds as they are in themselves. Other cultures do not ask about the universe as it is in itself, in the sense of modern Western science. They have no notion of a view from nowhere. Only because we have such a notion are we committed to the claim that our definition of natural kind terms captures the meaning of these terms for anyone correctly using them any time anywhere. Granted that our scientific understanding, if true, would be true in the world of the Egyptians even though they couldn’t understand it, it doesn’t follow, that what they meant by gold is determined by our science. Not that they had a rival universal account of gold as a natural kind. but neither were they deflationary realists avant la lettre. Gold being sacred presumably was not understood as relative to their description of it, but neither was gold’s sacredness understood as a universal truth about gold that all must acknowledge. Presumably, for them, gold being sacred revealed one aspect of it not necessarily revealed from any other perspective. This is presumably how the Homeric Greeks and the Native American tribes viewed the different gods worshipped by themselves and their neighbors. They presumably sensed that they neither discovered nor invented their classification of things and their gods, but drew on their form of life to reveal nature and its kinds from their own perspective. They thus implicitly took for granted that the way nature is revealed depends on what Heidegger calls “different kinds of seeing and questioning natural events.”46 The deflationary realist correctly believes that once world disclosing is recognized it would be mistaken to claim, as Kripke does, that the descriptions of natural kinds relied on in each particular world must either correspond to the structure of the universe as it is in itself, or be false.47 But it doesn’t follow that one has to give up a robust understanding of correspondence. Our scientific theory, if true, tells us what the property of gold that accounts for its other physical properties is, but this needn’t be the whole story. As Heidegger puts it: “The statements of physics are correct. By means of them, science represents something real, by which it is objectively controlled. But . . . science always encounters only what its kind of representation has admitted beforehand as an object possible for science.”48 On this view, the Egyptians’ understanding of the essential property of gold, if true, would also correspond to or reveal an aspect of nature. Given his understanding of supersession, Taylor would claim that we could, at least in principle, have taught the Ancient Egyptians our science, and with it the distinction between the in-itself and the for-us. They could

78

Hubert L. Dreyfus

then see both that gold is a natural kind in our sense with its essential property being an atomic number of 79, and that our disenchanted understanding of nature overlooks the fact that nature is sacred, and its kinds have a sacred essence our science can’t see. Even a view from nowhere of things as they are in themselves is only one limited way of disclosing them. Again, as Heidegger puts it, “What is represented by physics is indeed nature itself, but undeniably it is only nature as the object-area, whose objectness is first defined and determined through the refining that is characteristic of physics.”49 Thus, whereas gold’s physical property of being untarnishable is causally explained in terms of universal laws by our science and its view from nowhere, gold’s essential sacred property of shining with divine radiance may only be accessible to Egyptian religious practices.50 The kind of correspondence claim implicit in the practices of premodern cultures, if spelled out, would then amount to the claim that they have practices for gaining a perspective on reality that corresponds to one aspect of reality without claiming to have a view from nowhere that reveals objective reality as it is in itself. The aspect such practices revealed might have causal properties that could only be activated by those specific practices, and so would not be discoverable by a disenchanted science with a view from nowhere. Hence, what might seem a mystery or even an impossibility from the standpoint of our science might have a causal explanation within a given set of practices that reveal another type of causality. In the most extreme conceivable case, these culturally activated causal properties might even override the causal properties discovered by our science. If confirmed, repeatable levitation would be such a case and, our physics would then have to be revised to take account of such a phenomenon.51 All this leads to the conclusion that, although according to our disenchanted science it is true everywhere, whether or not anyone knows or cares about it, that gold has an atomic number of 79 since this property explains all the causal properties our science can see, it is only relative to our disenchanted way of questioning natural events that having an atomic weight of 79 is taken to be the essential property of gold. More generally, there is no single essential property of gold. Given the above considerations one has to be a plural realist where essences are concerned. Furthermore, although we don’t at present know of any alternative irreducible theories of nature as it is in itself, there is much we don’t understand, and there may be other ways of getting at universal causal properties that Western science can’t grasp. The success of acupuncture has so far resisted all attempts to understand it in terms of Western medicine, and we may simply have to accept two accounts of the body, one in terms of molecules

Taylor’s (Anti-) Epistemology

79

and electrical impulses, and another that plots the paths of a kind of energy that can’t be understood in terms of our current physics. We may also be seeing signs of a need for two independent accounts of reality, one describing those aspects of nature revealed to detached observors as it is in itself and another account of reality as it is revealed to involved human beings. Scientists and philosophers have, after all, so far failed to reconcile mechanical theories of physical reality with the seemingly undeniable facts of free will, consciousness, and meaning. Convergence in all these cases would certainly be satisfying and would reassure us that our theories describe an independent reality, but we have to leave open the possibly that there is no single privileged way the universe works. Because he has broken free of the last version of the inner/outer mediational picture – the claim that we must be imprisoned in our description of reality – Taylor can agree with Rorty that there is no one language for correctly describing the universe, while holding, contra Rorty, that there could well be many languages each correctly describing a different aspect of reality. Taylor’s anti-epistemology could then be characterized as pluralistic robust realism. Notes 1. Edmund Husserl, Cartesian Meditations: An Introduction to Phenomenology. Dorion Cairns (trans.), The Hague: Martinus Nijhoff, 1960. 2. Letter to Gibieuf of 19 January 1642; English version in Descartes: Philosophical Letters. Anthony Kenny (trans.), Oxford: Oxford University Press 1970, p. 123. 3. As Taylor has not yet published his detailed phenomenological alternative to current epistemology, which is tentatively entitled Retrieving Realism, I’ll be quoting at length from personal communications on the subject unless otherwise noted. 4. See Martin Heidegger, Being and Time. J. Macquarrie and E. Robinson (trans.), Oxford: Blackwell, 1962; Maurice Merleau-Ponty, Phenomenology of Perception. C. Smith (trans.), London: Routledge, 1962, and Samuel Todes, Body and World. Cambridge, MA: M.I.T. Press, 2001. 5. Martin Heidegger, Being and Time. J. Macquarrie and E. Robinson (trans.), Oxford: Blackwell, 1962, p. 405. 6. Charles Taylor, “Overcoming Epistemology” in Philosophical Arguments, Cambridge, MA: Harvard University Press, 1995, p. 12. Emphasis in the original. 7. Richard Rorty. Philosophy and the Mirror of Nature, Princeton: Princeton University Press, 1979, p. 178. 8. “A Coherence Theory of Truth and Knowledge” in Ernest Lepore (ed.), Truth and Interpretation: Perspectives on the Philosophy of Donald Davidson, Oxford: Blackwell, 1992, p. 310. 9. Ibid.

80

Hubert L. Dreyfus

10. Charles Taylor, “Foundationalism and the Inner-Outer Distinction” in Reading McDowell on Mind and World. Nicholas H. Smith (ed.), London: Routledge, 2002, p. 111. 11. If one takes justifying to be giving reasons for holding his belief, it does seem plausible to hold with philosophers like Davidson and Rorty that, in using his perceptual skill, Johnny would not be justifying his assertion that the picture was crooked, but, rather, Johnny has acquired the nonpropositional perceptual experience of the picture’s being askew, and the disposition to act and judge accordingly. What is at stake philosophically is whether or not there is another way of grounding a belief than basing it on another already-accepted belief. On the view Taylor is defending, beliefs concerning the perceived world are grounded when we achieve a maximal grip on the object in question. MerleauPonty calls this relation of support “motivation.” See Mark Wrathall, “Motives, Reasons and Causes,” Cambridge Companion to Merleau-Ponty. Taylor Carman (ed.), Cambridge, U.K.: Cambridge University Press, forthcoming. 12. John McDowell, Mind and World, Cambridge, MA: Harvard University Press, 1993. 13. “Foundationalism and the Inner-Outer Distinction” in Reading McDowell on Mind and World Nicholas H. Smith (ed.), London: Routledge, 2002, p. 113. 14. John McDowell, Mind and World, Cambridge, MA:. Harvard University Press, 1993, p. 61. McDowell’s description helps one understand why Merleau-Ponty calls this kind of grounding, “motivation.” His argument also runs very close to Taylor’s here, including his critical treatment of the same paper by Davidson, and even his citing the same quotations about the impossibility of getting beyond beliefs, and not being able to jump out of our skins. Ibid. Lecture I, sect 6. For Davidson, see “A Coherence Theory of Truth and Knowledge” in Ernest Lepore (ed.), Truth and Interpretation: Perspectives on the Philosophy of Donald Davidson, Oxford: Blackwell, 1992, pp. 307–19. 15. For instance, when he says things like the following: “But we really cannot understand the relations in virtue of which a judgment is warranted except as relations within the space of concepts: relations such as implication or probabilification, which hold between potential exercises of conceptual capacities.” John McDowell, Mind and World, Cambridge, MA: Harvard University Press, 1993, p. 7. The “exercises” referred to here surely must be propositional. 16. Ibid., p. 5. 17. Ibid., p. 24. 18. Ibid., p. 26. 19. Ibid., p. 32. 20. Maurice Merleau-Ponty, The Structure of Behavior. A. L. Fisher (trans.), Pittsburgh: Duquesne University Press, 1983. 21. John McDowell, Mind and World, Cambridge, MA: Harvard University Press, 1993, p. 66. 22. See Charles Taylor, The Explanation of Behavior, London: Routledge and Kegan Paul, l964, and “Overcoming Epistemology” in Philosophical Arguments, Cambridge, MA: Harvard University Press, 1995, pp. 1–19.

Taylor’s (Anti-) Epistemology

81

23. John Searle, Minds, Brains, and Science, London: Penguin, 1989, p. 21. 24. John Searle, Intentionality: An Essay in the Philosophy of Mind, Cambridge, U.K.: Cambridge University Press, 1983, p. 230. 25. As far as the issues here are concerned, we can ignore the question whether what is doing the processing is the brain or the whole nervous system. So, for the sake of simplicity, I’ll refer to the brain. 26. Phenomenological evidence could perhaps be used to argue that the action/ perception feedback loop between the agent and the world (what MerleauPonty calls the intentional arc) and the sense of relevance it generates is too ramified and open-ended to be simulated by a virtual reality program. Taylor’s holistic doubts on this point are certainly well taken but not relevant here. 27. Maurice Merleau-Ponty, Phenomenology of Perception. Christopher Smith (trans.), London: Routledge, 1962, p. 344. 28. Charles Taylor, “Foundationalism and the Inner-Outer distinction” in Reading McDowell on Mind and World Nicholas H. Smith (ed.), London: Routledge, 2002, p. 114. 29. Maurice Merleau-Ponty, Phenomenology of Perception. Christopher Smith (trans.), London: Routledge, 1962, p. 355. 30. Whether the beliefs of a brain in a vat would be false, and if so, what the implications for the contact realist are, is a separate and more complicated question that must await my discussion of Taylor’s account of truth. 31. Rorty insists that speaking of propositions corresponding to reality is a misleading way of speaking of our everyday activity in that it calls up a representationalist picture that pragmatists can and should leave behind. He says: “I do not think that either language or knowledge has anything to do with picturing, representing, or corresponding, and so I see formulating and verifying propositions as just a special case of what Taylor calls ‘dealing’ and I call ‘coping.’ “Charles Taylor on Truth,” Truth and Progress: Philosophical Papers, Vol. 3, Cambridge, U.K.: Cambridge University Press, 1998, p. 95. But, if Taylor is right, this assimilation of propositions to modes of coping begs precisely the question at the heart of the mediational issue and confronts Rorty with a dilemma. Either propositions, as opposed to coping practices, must in some sense be inner and represent the world in that they have content and one can separate their conditions of satisfaction from whether these conditions are satisfied, or beliefs don’t have propositional content, in which case, according to Rorty, they cannot provide rational justification for other beliefs. 32. We’ve already seen that Taylor’s rejection of mediational antirealism turns on his account of how the embodied agent both shapes the world and is open to what shows up in it. He tells us: We need (i) to allow for a kind of understanding which is pre-conceptual, on the basis of which concepts can be predicated of things. For this, we need (ii) to see this understanding as that of an engaged agent, determining the significances (sens, Sinne) of things from out of its aims, needs, purposes, desires, and (iii) to see that the most primordial and unavoidable significances of things are or are connected to those involved in our bodily existence in the world: that our field is shaped in terms of up or down, near or far, easily accessible or out of reach, graspable, avoidable, and

82

Hubert L. Dreyfus so on. This is where Merleau-Ponty’s contribution, enlarged and developed recently by Todes, has been so crucial.

Thus, according to Taylor, we have direct access to the everyday world precisely because it is organized by and for embodied beings like ourselves. 33. Richard Rorty, “Charles Taylor on Truth,” Truth and Progress: Philosophical Papers, Vol. 3, Cambridge, U.K.: Cambridge University Press, 1998, pp. 93–4. 34. Taylor sometimes seems to undermine any hope of saving himself from antirealism, as in the following comment in “Foundationalism and the Inner-Outer Distinction” in N. Smith (ed.), Reading McDowell on Mind and World, London: Routledge, 2002, p. 113. . . . I might lay out the environment I normally walk about in by drawing a map. But this wouldn’t end the embedding of reflective knowledge in ordinary coping. The map becomes useless, indeed ceases to be a map in any meaningful sense for me, unless I can use it to help me get around. Theoretical knowledge has to be situated in relation to everyday coping to be the knowledge that it is.

35. 36.

37. 38.

39.

40.

With the claim that theory, like a map, must be abstracted from everyday practices yet gets its intelligibility only through its relation to them, Taylor seems committed to a kind of “only through” mediational antirealism. Merleau-Ponty makes a similar claim about maps at the beginning of Phenomenology of Perception: “Every scientific schematization is an abstract and derivative sign-language, as is geography in relation the countryside.” Christopher Smith (trans.), London: Routledge, p. ix. And he draws a Rorty-like conclusion. “L’objectivit´e absolue et derni`ere est un rˆeve. . . . Nous ne pouvons pas nous flatter, dans la science, de parvenir . . . a` un objet pur de toute trace humaine.” Maurice Merleau-Ponty, Causeries 1948, Paris: Seuil: 2002, pp. 15–16. Richard Rorty, “Charles Taylor on Truth,” Truth and Progress: Philosophical Papers, Vol. 3, Cambridge, U.K.: Cambridge University Press, 1998, pp. 93–4. Samuel Todes, Body and World, Cambridge, MA: M.I.T. Press, 2001. One needn’t bring in gravity. We can simply point to whatever pulls everything, including us, down; something we have to get aligned with in order to cope successfully. See Piotr Hoffman’s introduction to Body and World on the use Todes makes of his observations concerning the vertical field of influence. Maurice Merleau-Ponty, Phenomenology of Perception. Christopher Smith (trans.), London: Routledge, p. 322. We need to add “normally” here because, if Taylor admits the conceivability of the brain-in-the-vat, as I think he should, it might turn out that there is no Cosmos. But that is not an objection to Taylor’s claim that prima facie we experience having to get in sync with something independent of us that both defeats and supports our activity and that, in so doing, we reveal its structure. More on the status of the truth claims based on this experience in Section VII. Martin Heidegger, “The Age of the World Picture,” The Question Concerning Technology and Other Essays. W. Lovitt (trans.), New York: Harper Torchbooks, 1977. Richard Rorty, “Charles Taylor on Truth,” Truth and Progress: Philosophical Papers, Vol. 3, Cambridge, U.K.: Cambridge University Press, 1998, p. 86.

Taylor’s (Anti-) Epistemology

83

41. Ibid., pp. 86–7. 42. In science the new context that replaces the world is the theory itself. As Heidegger points out, and Taylor agrees, scientific facts are always theory laden; they are, however, deworlded, that is, disconnected from our purposes. 43. Charles Taylor, “Foundationalism and the Inner-Outer Distinction” in Reading McDowell on Mind and World. Nicholas H. Smith (ed.), London: Routledge, 2002, p. 115. 44. Maurice Merleau-Ponty. Sense and Non-Sense. H. L. Dreyfus and P. Allen Dreyfus (trans.), Evanston, IL: Northwestern University Press, 1964. 45. See Hubert L. Dreyfus, “Comments on Cristina Lafont’s Interpretation of Being and Time,” Inquiry. A. Hannay (ed.), Vol. 45, No. 2, June 2002. Saul A. Kripke, Naming and Necessity, Cambridge, MA: Harvard University Press, 1972, p. 132. 46. Martin Heidegger, “The Age of the World Picture” in The Question Concerning Technology and Other Essays. W. Lovitt (trans.), New York: Harper Torchbooks, 1977, p. 117. 47. That doesn’t mean that what a culture’s natural kind terms mean is up for grabs. Rorty had it when he spoke of a culture’s fundamental vocabulary. A culture’s way of understanding nature, for example, is too bound up with its other practices to be changed at will. See Richard Rorty, Contingency, Irony, Solidarity, Cambridge, U.K.: Cambridge University Press, 1989. 48. Martin Heidegger, “The Thing” in Poetry, Language, Thought. A. Hofstadter (trans.), New York: Harper and Row, 1971, p. 170. Emphasis in the original. 49. Martin Heidegger, “Science and Reflection” in The Question Concerning Technology and Other Essays. W. Lovitt (trans.), New York: Harper Torchbooks, 1977, pp. 173–4. 50. It helps to remember Wittgenstein’s point in his critique of Fraser’s The Golden Bough, that we should understand a rain dance which is normally performed just before the rainy season, not as a mistaken attempt to cause rain in our sense of cause, but as a celebration in advance of the hoped for rain. As Wittgenstein points out: “[T]oward morning, when the sun is about to rise, rites of daybreak are celebrated by the people, but not during the night, when they simply burn lamps.” Ludwig Wittgenstein, Philosophical Occasions – 1912–1951. James C. Klagge and Alfred Nordmann (eds.), Indianapolis and Cambridge, U.K.: Hackett Publishing Company, 1993, p. 137. 51. There are also privileged perspectives on reality that don’t purport to be a view from nowhere but purport to supersede all other views. They range from Plato’s claim that the philosopher has such a perspective once he or she emerges from the cave, to religions such as later Judaism, Buddhism, and Christianity, that each claims to have a true and universally valid perspective, but has no notion of a view from nowhere.